Sep 27, 2010

Going Freemium: One Year Later

On September 1st, 2009 we announced that MailChimp was going freemium. On that day, we had 85,000 users. Now, slightly more than a year later, we have more than 450,000 users. We grew our user base five times in one year.

Earlier this month, we actually doubled our freemium plan from 500 subscribers to 1,000 subscribers. So now, even more people can take advantage of MailChimp’s powerful email marketing and social features. We had been averaging around 30,000 new users per month (about 1,000 per day), but since we increased the freemium plan this month, we’re seeing +2,000 new user days.

Another thing that’s increased dramatically since going freemium is the number of lunches I’m invited to; seems entrepreneurs and VCs really want to "pick my brain" about how freemium is doing for us. Usually, it’s because they think freemium might be that silver bullet they’ve been searching for. It can be, but you’ve really gotta be careful not to point that bullet at yourself…

See, I’ve heard a lot of discussion about MailChimp’s success with freemium, but it’s scary that people would want to model their companies around us. I say it’s scary, because they don’t know our history, and they don’t know our motivation. I came across this article recently: Why Free Plans Don’t Work. They actually mention MailChimp in a positive way, but the discussion in the comments show that people really need a little insight into what drove us to go freemium. So here goes…

First, the stats

Yeah, I know you’re just here for the numbers. So here ya go:

  • We are adding over 30,000 new users and 4,000 new paying customers each month.
  • Since starting freemium 12 months ago, the number of paying customers has increased over 150% and profit has increased over 650%.
  • Our profit has increased mainly because our cost of acquisition has been dropping. It decreased 8% in just the last quarter to under $100.
  • We’re delivering roughly 700 million emails per month

"Yeah, but they’re all tiny customers, right?"

Something that might surprise people is that going freemium didn’t totally skew our user base to teeny-tiny accounts full of "extremely small businesses" with extremely small lists. We’ve actually seen our number of large customers grow. A lot.

In April of 2010 (earlier this year), we ran some numbers and found 12% of our paying customers had lists greater than 10,000 subscribers.


12% of our users had lists greater than 10k subscribers

But at the end of August 2010, that 12% grew to 20%:


20% of our users have lists greater than 10k subscribers

In that same period, the percentage of our total revenue from those larger customers grew from 48% to 65%. I guess that’s to be expected when you make your lower tiers free. It’s also mathematically predictable that over time, our average revenue per user (ARPU) will decrease, while our competition’s will increase (assuming we’re getting the lower tiers). We’ll have to see if they start touting higher ARPUs, I suppose.

Another way to look at this is the amount of email we’re sending every month for customers with lists over 50,000 subscribers:


Roughly 200 million emails per month for high-volume customers

This first bump, in June of 2008, was after we launched MailChimp v3. The next spike was September 2009, when we launched freemium. We’re now sending roughly 200 million emails per month for these larger customers.

The fact that we’re attracting larger users, with larger lists, and with more advanced needs, would seem counter-intuitive to most people.

Especially to some of our competitors, who like to say, "You get what you pay for."

We think we’ve used technology and innovation (much like Google) to flip that silly old argument on its head. With MailChimp, you get tons of powerful features, and awesome deliverability, for free. And people aren’t leaving as their lists grow. They’re happy to stay and grow with us, and take advantage of the new, innovative features we’re always launching.

Why on earth could going freemium bring in these larger and larger paying customers?

Because we did everything totally bass-ackwards.

MailChimp’s History

If you’re trying to understand our approach to freemium, you need to know a little more about our company history. First, there are some things people assume about MailChimp that are simply not true. For one, we are not a startup. We have the benefit of many, many years of making — and learning from — tons of mistakes. It also helped that we started our company in the midst of the dot-com meltdown (see: Why to Start a Startup in a Bad Economy, by Paul Graham). Some people think we’re a startup because they just didn’t know about us until we launched v3 in May of 2008. Now, they see our ads popping up everywhere, and they think we’re a startup like them. And when they tell me they’re thinking of going freemium because they’re really tired of waiting for that "needle to move," I tell them to relax because they’ve got about 10 more years to go. They never seem to like my advice.

So take it from this guy instead:



The second thing people don’t always get about us is that we’re not taking the conventional, "build a stripped down product, make it free, then up-charge if people want more features" approach to freemium. We spent years building up a powerful, affordable, profitable, self-serve product. We invested heavily in our API, which now has more than 70,000 users. We got smarter at deliverability, scalability, and abuse prevention. Then, the cloud made all of the above even cheaper. We took advantage of those savings and made stuff free.

Years of Pricing Experimentation
Here’s another piece of history. Ever since inception, I’ve been fascinated with the art and science of pricing. I’ve tinkered with pay-as-you-go and monthly plans for $9, $9.99, $25, $49, $99.99 and so on. We’ve changed our pricing models at least a half-dozen times throughout the years, and along the way we tracked profitability, changes in order volume, how many people downgraded when we reduced prices, how many refunds were given, etc. We’re sitting on tons of pricing data. When we launched our freemium plan in 2009, you betcha we used that data to see what would happen if we cannibalized our $15 plan.

If we had started with freemium at ground zero, the story would’ve been different. Here’s what I mean…

The 10:1 Ratio

A lot of startups who are starting at ground zero (no customers) focus on the conversion rate of freemium-to-paid, so that’s the first question they usually ask me. It’s smart to worry about that, because frankly, it sucks. Matt Brezina from Xobni did a presentation at the Freemium Summit early this year that basically made the point that freemium isn’t a new concept; it’s just a new word. The concept’s been around forever, and is used by your local art museum, the park, the government, and in other places you’ve probably never thought about. So Mr. Brezina studied their conversion rates. The key slide in his presentation, imho, was slide 6:

10:1 - the ratio of free users to paid

10:1 - the ratio of free users to paid, from a Freeemium study presented by Matt Brezina of Xobni

Throughout history, and across all the businesses he researched, the ratio of free-to-paid-subscribers ultimately ended up at the dismal ratio of 10:1. There were a lot of awesome presentations at that Freemium Summit. But this was the presentation (just this one slide, really) that stuck in my brain.

Look at that ratio. For every one paying customer, you’ll have ten free users.

Is Freemium Right for You?

So the question to ask yourself is whether or not your "one" is big enough to pay your bills yet. For eight years, our company never thought about freemium. We didn’t even know the concept existed. For eight loooong years, we were focused on nothing but growing profits. If you had brought up the concept of "freemium" with us during those eight years, we probably would’ve looked at you like you were eff’ing insane, then went back to work. In fact, when we launched MailChimp in 2001, we didn’t even have a free trial option. It took us several months to get that in place (I think that was more a function of my co-founder and I getting laid off from a dot-com than anything else). And when we launched our free trial, it only allowed something ridiculously low, like 25 emails. In other words, we’ve been laser-focused on the "1" side of that 10:1 ratio. We’d never consider freemium until our "1" was big enough. Enough to pay for 70+ employees, their health benefits, stash some cash for the future, etc.

I think there are too many startups out there who are interested in going freemium because that big "10" number is so attractive. This is dangerous when they don’t even have the "1" yet. How will they pay their bills while they figure out how to "monetize?" Answer: they will need to borrow that money. Does your VC have the patience for the long term, while you try to figure out how to "monetize" and build up that measely "1" number? Answer: No — no they don’t. Build up that "1" before you chase the "10." After you’ve got your "1" all set, use VCs to help you chase after that "10" (if you must). That’s my personal opinion. Disclaimer: I’m wrong about 99% of the time.

A side benefit of taking our bass-ackwards approach to freemium is that we don’t think of our free users as pesky, bandwidth-hogging "freeloaders" that we have to monetize in some way. We love them just as much as the people who pay us money. Because we have the data that shows they will pay us money.

Our Real Motivation: Free Ice Cream!

I was never an ice cream snob, until I met my friend Mark in college. When I was a kid, Breyers was the most premium ice cream that ever showed up in our freezer, and if I was really lucky, some friend of a friend from school would invite me to a birthday party at a Baskin Robbins. One day, Mark introduced me to "Ben & Jerry’s free ice cream day:"


I had no idea what the heck Ben & Jerry’s was. He was appalled, and had to educate me. On our way there, he went on and on about how awesome Ben & Jerry’s was, what a great brand it was, and what a great experience the whole thing was for him (I’m pretty sure he mentioned the Grateful Dead, Phish, hippies hugging trees, world peace, and all the other quirky things associated with Ben & Jerrys at that time).

So he takes me to free cone day, and it’s this small ice cream shop in an upscale, yet hippy part of town. There was actually a line coming out of the store. I remembered thinking, "This never happened at Baskin Robbins." When we finally got to the front of the line, my friend walked me through all the different flavors, and had me taste samples of every single one of them. While I was just trying to find Rocky Road (heh), he was busy buying up pints of different flavors to take home. He had totally "drunk the kool aid" as they say. And I gotta admit, I was drinking it too. I walked out of there with a cone of Chunky Monkey (of course) and have been a fan of Ben & Jerry’s ever since.

That, in a nutshell, was my inspiration and motivation in offering the freemium program at MailChimp.

We’ve worked hard for many years to build a powerful service that actually makes email marketing fun. And I want the entire world to experience that fun. It’s also kind of a cool idea to think of bajillions of "serious" business newsletters being distributed with little monkeys in their footer. And it was also a way to thank our customers. Many of them were able to switch from paid accounts to free. And as we double our freemium plan, even more will do the same (even if it does "drive them a little nuts"). Speaking of thanking our customers, we’re also giving away 5,000 MailChimp t-shirts inside the app. Send an email, and boosh — you might win an awesome t-shirt. And did I mention we’ve made Social Pro free until March 2011?

I’ve worked in the stuffy corporate world. It was an extremely valuable learning experience, but it was so — stuffy. It needs more monkey.

Of course, when I presented my "Let’s share MailChimp with the world and make it a monkey-er place" master plan with my co-founder, our COO, and lead engineer, they put a little bit more serious thought into the financial impact, our deliverability impact, abuse vectors, staffing requirements, and on and on. I don’t want anybody to think we launched freemium without a lot of careful thought and planning. And ever since launching, we’ve monitored our stats, and learned more about our freemium users, their impact on our business, and ultimately, used that data to decide whether or not we should double it this year.

There’s a lot of number crunching and analysis that goes along with freemium. But if anybody out there thinks this was motivated by some kind of competitive business or marketing or pricing strategy, I’m sorry to disappoint you: we just did it for fun. If it’s not fun, what’s the point, right?